Influence of skin testing and blood storage on interferon-γ production in cattle affected naturally with Mycobacterium bovis

BOVINE tuberculosis (TB) is a disease of economic and zoonotic importance caused by the bacterial pathogen Mycobacterium bovis. In the UK, infected cattle are identified using the single intradermal comparative tuberculin test (SICTT). Recent studies have demonstrated that the parallel use of a commercial test that detects antigen-stimulated bovine interferon-γ (IFN-γ) in whole blood cultures, the Bovigam assay (Prionics) (Wood and others 1991), can improve the sensitivity for detecting infected animals compared with the use of either of the tests used in isolation (Ryan and others 2000, Collins 2002). The Bovigam IFN-γ ELISA is approved for use in bovine TB control programmes in the European Union, Australia, New Zealand and the USA. Factors that can influence the strength of IFN-γ responses include the delay between taking blood and setting up the cultures, and the interval between sampling and a recent skin test (Rothel and others 1992, Ryan and others 2000, Whipple and others 2001). In a study using cattle experimentally infected with M bovis, Whelan and others (2004) reported a reduction in the strength of IFN-γ responses if blood samples were stored overnight before testing, and also if samples were collected and tested on the day that skin-test reactions were read. These factors did not alter the diagnostic outcome of the IFN-γ test when the comparative use of bovine and avian tuberculins as test antigens was applied. However, a reduction in test sensitivity was observed when a cocktail of peptides derived from ESAT-6 and CFP-10 was applied (Whelan and others 2004); these antigens have been shown to be useful in overcoming the limitations of specificity of purified protein derivative (PPD) (Pollock and Andersen 1997, Vordermeier and others 1999, 2001). A potential limitation of the previous study was that immune responses in experimentally infected cattle are highly synchronised and may not reflect fully the range of responses in naturally infected animals. This short communication describes a study designed to address this issue, by investigating the effects of a recent skin test and the overnight storage of blood on IFN-γ responses in blood from cattle naturally infected with M bovis. Naturally infected tuberculin reactor cattle (bovine PPD [PPD-B] – avian PPD [PPD-A] >4 mm) were recruited to the study during routine UK SICTT surveillance operations. They were housed in secure accommodation and were repeat skintested up to three times, at 60-day intervals. Blood samples were collected not more than seven days before each skin test, which was performed on day 0 and again on days 3 and 10 after the skin test. Whole blood cultures were set up on the day of blood collection (fresh blood cultures) and also following the overnight storage of blood at ambient room temperature (24-hour blood cultures). Whole blood (250 μl) was cultured in duplicate in the presence of 10 μg/ml of PPDA and PPD-B (VLA – Weybridge), or a cocktail of 5 μg/ml of each of 21 overlapping 16-mer peptides mapping the entire sequences of ESAT-6 and CFP-10 (Louisiana State University Health Services Center), or in the absence of antigen. Plasma supernatants were harvested following 20 hours’ incubation at 37°C in 5 per cent carbon dioxide, and tested for IFN-γ using the Bovigam ELISA kit. The infection status of all the animals studied was confirmed by the culture of M bovis (data not shown). Fig 1 shows the IFN-γ responses of the PPDand peptide cocktail-stimulated cultures. In comparison to the responses of blood samples taken before the skin test (day 0 samples), no significant differences were observed for the day 3 or day 10 responses stimulated by PPD or peptide antigens, either for the fresh blood or 24-hour blood cultures (two-tailed paired t test), although a trend was observed towards stronger IFN-γ responses on day 3 than day 0 in all cases, a finding that was also reported by Whipple and others (2001). The measurement of IFN-γ responses in fresh blood cultures at days 24 to 31, 38 to 45 and 59 after the skin test also demonstrated no significant differences from those seen at day 0 (data not shown). This observation is consistent with the findings of Gormley and others (2004), in a similar study of naturally infected cattle in the Republic of Ireland. The absence of a suppression of IFN-γ production at day 3 is in contrast to previous observations in experimentally infected cattle (Whelan and others 2004). This may reflect differences in the stage or progression of disease in these animals, since the experimentally infected cattle had more extensive pulmonary pathology compared with the naturally infected reactors in this study (data not shown). Whipple and others (2001) reported a boost in the IFN-γ response following the application of a caudal fold test, and that IFN-γ responses remained elevated for up to 59 days. Similarly, Ryan and others (2000) also reported elevated responses to the Bovigam test eight to 28 days after a caudal fold test. These observations contrast with the present findings and those of Gormley and others (2004); one possible reason for this is that the caudal fold test may be more effective at boosting IFN-γ responses than the SICTT. Overnight storage of the blood did result in a lower response (Fig 1), but this had little influence on the diagnostic outcome of the assay when applying the comparative use of PPD as test antigens (Table 1). Studies in New Zealand and the USA have concluded that approximately equal proportions of animals could be defined as positive when the IFN-γ assay was performed either on the day of sampling or 24 hours after sampling, despite a similar drop in signal strength (Rothel and others 1992, Whipple and others 2001). Those results were obtained in reactor animals with confirmed TB (Rothel and others 1992) or with experimentally sensitised steers (Whipple and others 2001), and formed the basis of the New Zealand and USA policy of starting the assay on the day after sampling (within 24 hours). It is interesting to note that the PPD responder frequency for the 24-hour blood was actually higher than for fresh blood, due to the resolution of a PPD-A bias for some of the cattle (Table 1). An increase in specificity following overnight storage has also been reported by Ryan and others (2000). In contrast to these findings, which suggest that overnight storage of blood has no significant impact on the sensitivity of the IFN-γ test, Gormley and others (2004) reported that the delay in culturing the blood reduced the test sensitivity significantly. In that study, three groups of IFN-γ test-positive animals, namely, SICTT-reactors with confirmed TB (group 1), a group of SICTT-positive animals from surveillance operations (group 2) six of 29 of which had confirmed TB and a group of 60 SICTT-negative animals on which no postmortem examinations were performed (group 3), were assessed. Applying the Bovigam test to the group 1 animals within eight or 24 hours did not decrease the sensitivity of the test significantly. However, in the other Veterinary Record (2007) 160, 660-662